Acute myelogenous leukemia (AML) involves the expansion of a clone of self-renewing progenitor cells which fail to differentiate completely and a disease course in which the leukemic cells eventually develop resistance to chemotherapeutic agents. The underlying concept of this research is that common genetic mechanisms may contribute to the escape of leukemic cells from both physiologic and chemotherapeutic control. The experimental strategy is to select sub-lines of the human promyelocytic leukemia cell line HL-60 for resistance to chemically-unrelated inducers of differentiation using multi-step selection procedures reported in other systems to produce amplification of genes relevant to the resistance mechanism. Sublines resistant to 6-thioguanine (6TG[unreadable]R[unreadable]) and retinoic acid (RA[unreadable]R[unreadable]) acquired numerous double minutes (dm) which are reliable indicators of amplified genes. Studies with the 6TG[unreadable]R[unreadable] sub-line indicate that presumptive amplified genes are associated not with cytotoxic-resistance but with cytodifferentiative-resistance, although the mechanism is unknown. Four inter-related approaches are considered: (1) to partially-purify the dm in order to prepare a DNA library highly-enriched in amplified sequences; (2) to identify amplified fragments in restriction endonuclease digest of variant cell DNA, using reversed phase chromatography followed by agarose gel electrophoresis; (3) to prepare Cot-selected, nick-translated DNA probes from resistant (res) and wild-type (w-t) DNA and, after "clean-up" by hybridization to w-t DNA, to differentially-screen res DA, e.g., from (1) or (2) or res polysomal RNA, for homologous amplified sequences; and (4) to transmit a phenotypic property of the res DNA to an appropriate target cell. Our preliminary tests suggest that the initial combined use of approaches (2) and (3) will be most efficacious and conclusive. The long-range objectives are to determine the genetic and molecular mechanisms of cytodifferentiative resistance to 6TG and RA, to determine if this model may be useful clinically for detecting genetic evidence of drug-resistance, and to determine if this mechanism involves amplification of genes related to the differentiation process itself as perturbed in AML. (M)